Google reveals its plans to use machine learning (ML) to make web browsing less painful. According to the company, machine learning is already used in Chrome to allow features such as indicating likely malicious websites and grouping browsing history based on particular topics. However, it intends to introduce new features informed by machine learning in the future.
It begins with disturbing permission requests from websites. Almost everyone would have experienced such disturbances before – every website when visited seeks permission for sending notifications, knowing your location, or obtaining access to something on your device.
Chrome forecasts when permission prompts are unlikely to be allowed based on the way, the user previously cooperated with similar permission prompts, and mutes these undesired prompts to assist people to browse the web with minimal interruption, Google says, and it will also be releasing an ML model that makes these forecasts entirely on-device in the next version of Chrome.
This feature shows how Google can reactively use machine learning. On the ardent side, the company claims that Chrome’s future versions will make use of the technology to enable a toolbar that can automatically adapt to user behavior.
Google states that it wants to ensure that Chrome meets you where you’re at, so shortly, they will be using ML to adjust the toolbar in real-time – emphasizing the action that’s most helpful at that moment (e.g., share a link, voice search, and so on). (According to the company, Chrome users will still be able to manually customize the toolbar, so this feature shouldn’t be overly disruptive.)
Source link
